Limb Salvage Using Original Low Heat-Treated Tumor-Bearing Bone

Abstract

Limb salvage, using original low heat-treated tumor-bearing bone and a conventional joint prosthesis, was done in six patients with malignant tumors of the proximal humerus (one patient with chondrosarcoma and five patients with osteosarcoma) and in six patients with tumor of the proximal femur (two patients with malignant spindle cell sarcoma and four patients with osteosarcoma). Wide excision of the lesion was done and the tumor and surrounding soft tissues were removed. The excised bone was treated with heat and the prosthesis was inserted into the treated bone and fixed with cement. This construct was reinserted into the original site and anchored to the host bone with a plate. The overall union rate of the low heat-treated bone with normal host bone was 91.7%, and the mean union time was 4.6 months (range, 3–7 months) after surgery. The functional result of the proximal femur and proximal humerus were 76.7% and 56.8%, respectively, using the Musculoskeletal Tumor Society functional evaluation system. Complications included hip dislocation in one patient, fracture of the low heat-treated bone in two patients, and absorption of the low heat-treated bone of the humerus in four of six patients. The 5-year survival rate of the low heat-treated tumor-bearing bone was 83.3% using Kaplan-Meier survival analysis. Based on the results of this study, limb salvage using original low heat-treated tumor-bearing bone seems to be effective in treating primary bone sarcoma with high survival and acceptable complication rates, circumventing the complications of allograft bone.

Limb salvage of primary bone sarcoma has become the standard surgical treatment with the introduction of newly-developed adjuvant chemotherapy, preoperative chemotherapy in particular. ^1,5,13,15,20 There are several options for limb salvage: endoprosthesis, ^9,10,22,24 allograft bone, ^12,18,21,27 and original autoclaved bone with tumor, ^2–4,11,16 and each has its own advantages and disadvantages.

Limb salvage using an endoprosthesis is technically easy and saves operation time, but endoprostheses are expensive, frequently require a tailor-made design for each patient, and lack longevity. Limb salvage using allograft bone also is technically easy, but carries risks of disease transmission (autoimmune deficiency syndrome and hepatitis), immune reaction, nonunion, and absorption. Availability of allograft bone also is limited in East Asian countries because of cultural traditions. The original autoclaved bone with tumor is immunologically ideal because it uses the patient’s own bone. However, high temperature degrades proteins, including bone morphogenetic protein; therefore, nonunion and absorption of autoclaved bone frequently occur.

In the current study, another option is presented for limb salvage using original low heat-treated tumor-bearing bone and a conventional joint prosthesis. Wide excision of the tumor, removal of the muscle and periosteum from the excised bone, and curettage of the tumor are done. The bone is treated with low heat and a conventional prosthesis is inserted into the excised bone segment. Firm fixation is done with cement, and this construct is reinserted into the joint and fixed to the host bone with a plate. The outcomes of this procedure are evaluated to determine the usefulness of this option.

MATERIALS AND METHODS

This study included 12 patients who had wide excision and limb salvage with original low heat-treated tumor-bearing bone and a conventional prosthesis from 1994 to 1998. The selection criteria for this procedure were patients with a malignant bone tumor in the proximal femur or humerus requiring reconstruction of the hip or shoulder, with relatively intact cortical continuity of the proximal femur or humerus, and without distant metastases. The mean followup was 31.8 months (range, 11–62 months). Seven of the 12 patients were male and five were female, and the mean age at surgery was 26.1 years (range, 10–58 years).

Pathologically confirmed diagnoses were osteosarcoma in nine patients (four of the proximal femur and five of the proximal humerus), malignant spindle cell sarcoma in two patients (proximal femur), and chondrosarcoma in one patient (proximal humerus). Based on the system of Enneking et al, ⁸ the surgical stagings were IIA in three patients and IIB in nine patients (Table 1).

TABLE 1: Patients’ Data

Plain radiography of the lesion site, whole body bone scans, magnetic resonance imaging (MRI) of the lesion site, plain radiography of the chest, and high resolution computed tomography (CT) of the chest were done for diagnosis and tumor staging. An open biopsy was done in all patients, and a pathologic diagnosis was made. Three cycles of preoperative chemotherapy with intraarterial infusion of cisplatin (120 mg/m²) and intravenous doxorubicin (25 mg/m²) were administered in nine patients. Preoperative chemotherapy was not administered in the remaining three patients who had malignant spindle cell sarcoma and chondrosarcoma.

After three cycles of preoperative chemotherapy, a wide excision of the lesion was done. The distal and proximal margins of the wide excision were determined as 3 cm distal and proximal to the gadolinium-enhanced lesion of the femur or humerus as seen on preoperative MRI. In three of the patients with a lesion in the proximal femur, an extraarticular wide excision including the acetabulum was done, because the preoperative MRI showed high signal intensity in the joint cavity.

After the wide excision of bone and soft tissue, the soft tissue, muscle, periosteum, and tumor were removed from the excised bone. The neck of the femur or the humerus was cut and curettage was done in the medullary cavity. The excised bone was treated in a sterile saline bath at 65°C for 30 minutes. During this time, the soft tissue and the curetted specimen obtained from the distal margin of the host bone had frozen sections examined to confirm the surgical margin.

After the heat treatment, curettage of the medullary cavity was done again to adjust the prosthesis. The medullary cavity of the excised bone was filled with bone cement (DePuy CMW 1 original bone cement, Warsaw, IN) and the stem of a conventional joint replacement prosthesis (DePuy global shoulder system or Depuy Elite Plus hip system) was inserted. This construct then was fixed to the host bone (the normal distal portion of the femur or humerus) with a dynamic compression plate and cortical screws (Figs 1, 2). A Bard mesh (Davol Inc, Cranston, RI) that was wrapped around the inserted bone was used to suture the muscles, and tendons of the rotator cuff of the shoulder and abductor of the hip. The head and neck of the femur or humerus, and the first and second curetted specimens were examined histologically to determine tumor response to preoperative chemotherapy.

Fig 1A–D.: A 13-year-old girl had osteoblastic osteosarcoma of the femoral head and neck. (A) A T2-weighted coronal MRI scan shows invasion of the hip by a tumor mass (white arrow). (B) A preoperative plain radiograph shows a radiolucent lesion (arrow) on the inferomedial aspect of the femoral head and neck. (C) A postoperative radiograph shows the reconstruction of the hip with total hip arthroplasty after extraarticular excision, and the osteotomy site (white arrow). (D) A plain radiograph obtained at the 42-month followup shows a solid bony union (small white arrow). New bone formation was seen on the lateral aspect of the low heat-treated bone (large white arrow).

Fig 2A–D.: A 24-year-old woman had osteoblastic osteosarcoma of the proximal humerus. (A) A T1-weighted MRI scan with enhancement (arrow), and (B) preoperative plain radiograph show a radiolucent lesion (arrow). (C) A postoperative plain radiograph shows the reconstruction of the shoulder using a Neer hemiarthroplasty. (D) The plain radiograph obtained at the 53-month followup shows a solid bony union.

After surgery, three cycles of postoperative chemotherapy were administered. The chemotherapy regimen was determined depending on the rate of tumor necrosis of the head and neck and curetted specimens compared with that of the biopsy specimen. Postchemotherapy necrosis was determined histologically by the drop out of neoplastic cells with residual intact tumor matrix of varying morphologic features in most cases of osteoblastic type of osteosarcoma. In some cases, tumor necrosis was accompanied by an ingrowth of loose granulation tissue or fibrosis. Degenerative features of neoplastic cells, including no distinctly stained tumor nuclei, were histologic criteria for ascribing postchemotherapy necrosis; most of the chondroblastic osteosarcoma showed this pattern of postchemotherapy necrosis.

Tumor necrosis less than 90% was an indication to change the regimen to a high dose of methotrexate (10 g/m²), ifosfamide (1.5 g/m²), and etoposide (100 mg/m²). The patients were followed up postoperatively every month with plain radiographs of the lesion. Plain film analysis was used to evaluate the fusion status. The absence of an osteotomized line or the presence of callus formation crossing the osteotomized line was considered a union. Plain radiographs of the chest every month and CT scans of the chest every 3 months were obtained for detection of the lung metastasis. Serum alkaline phosphatase and lactic dehydrogenase every month and whole body bone scans every 3 months were evaluated for local recurrence and systemic metastasis.

Functional outcome on the last visit was assessed by a system devised by Enneking et al. ⁷ Statistical analysis was done using the Mann-Whitney U test, chi square test, Fisher’s exact test, and Kaplan-Meier survival analysis.

RESULTS

Union

The overall union rate of the original low heat-treated tumor-bearing bone with the normal host bone was 91.7% (11 of 12 patients). There was only one patient with nonunion. The union rate of the patients with proximal femoral lesions was 100% and that of the patients with proximal humeral lesions was 83.3%. The overall mean time for complete union was 4.6 months (range, 3–7 months) after surgery. The mean time for union was 4.2 months in proximal femoral lesions and 5.2 months in proximal humeral lesions. There were no statistically significant differences between the two groups for union rate (p = 1.000) and union time (p = 0.300) (Table 1).

Survival of Original Low Heat-Treated Tumor-Bearing Bone

There were two cases where the original low heat-treated tumor-bearing bone was fractured and there was one case of marked absorption with nonunion, which required revision surgery. Based on a Kaplan-Meier survival analysis, the mean survival time of the original low heat-treated tumor-bearing bone was 57.3 months, the median survival time was 62 months, and 5-year survival rate was 83.3%.

Complications

The overall complication rate was 33.3% (four of 12 patients). The complication rate for patients requiring revision surgery was 25.0% (three of 12 patients). There was one patient with a hip dislocation 3 months after a total hip arthroplasty, and closed reduction was done successfully. There were two fractures of the low heat-treated tumor-bearing bone 7 months and 11 months after surgery, respectively. Both were caused by slipping accidents. One patient had osteosarcoma of the proximal humerus and the other had osteosarcoma of the proximal femur. Internal fixation was done with plate and screws. Absorption of the low heat-treated tumor-bearing bone occurred in four of the six patients with proximal humeral lesions (Fig 1). Among those patients, one had severe absorption of the bone and nonunion, therefore, revision surgery was done. However, there was no absorption of the reinserted low heat-treated femoral bone. There were no superficial or deep wound infections.

Functional Outcome

The overall mean functional outcome was 66.5% (19.6 points). The mean functional outcome of the lower extremities of the patients with proximal femoral lesions was 76.7% (22.8 points), and that of the upper extremities of the patients with proximal humeral lesions was 56.8% (17.0 points). There were statistically significant differences between the two groups for functional outcome (p = 0.011) assessed by the Mann-Whitney U test and Fisher’s exact test (Table 1).

Oncologic Outcome

Four patients had 100% tumor necrosis, two patients had more than 90% tumor necrosis, and three patients had less than 90% tumor necrosis (80%, 70%, and 10% each) (Table 1). There was no local recurrence. There were three distant metastases, all to the lung. Two of the three patients who had a poor response to preoperative chemotherapy (less than 90% tumor necrosis) had metastases to the lung 17 months after surgery. However, one of the six patients who had a good response to preoperative chemotherapy had metastasis to the lung 34 months after surgery.

The mean time at which a metastasis was found was 22.7 months (range, 17–34 months) after surgery (Table 1). The overall mean survival was 52.4 months, the median survival was 52 months, and 5-year survival rate was 43.8%, based on a Kaplan-Meier survival analysis. The mean survival of the patients who had a poor response to preoperative chemotherapy was 36 months, the median survival was 17 months, and 5-year survival rate was 0%. All patients who had a good response were alive; therefore, a survival analysis was not done (Table 1).

DISCUSSION

The options for reimplantation of extracorporeally devitalized bone segments for reconstruction are autoclave, ^2–4,11,16 high-dose irradiation, ² microwave heat, ¹⁶ and low heat treatment. ^11,14 In 1956, Thompson and Steggal ²⁶ reported reimplantation of original autoclaved bone with tumor as a possible treatment of chondrosarcoma of the proximal femur. Resected bone was treated with high heat (131°C) for 5 minutes, a method extremely effective for tumor cell necrosis. However, high temperature degrades proteins, and degradation of bone morphogenetic protein occurs within 1 hour at a temperature above 70°C. ¹⁹ Also, the mechanical strength of the autoclaved bone decreases. ¹⁷ Therefore, delayed union, nonunion, or fracture frequently occurs after reimplantation of autoclaved bone. Lu et al ¹⁶ described an intraoperative in situ microwave heating technique which had good oncologic and functional results. The tumor bone was held at 50°C for 15 minutes.

Tumor cells, bacteria, and virus can be killed at 65°C. ¹⁰ Liebergall et al ¹⁴ reported that bone cells died completely after 30 minutes at 60°C. They concluded that the ability of controlled hyperthermia to eradicate viable cells without significant damage to the mechanical properties might have clinical relevance in limb salvage tumor surgery. Bohm and Stihler ⁴ did various low heat treatment studies and found that treatment at 45°C for 11 minutes was not sufficient to kill all tumor cells.

In the current study, low heat treatment was done at 65°C for 30 minutes. The union rate between the reimplanted original low heat-treated tumor-bearing bone to host bone was 91.7%, which was higher than comparable previously reported rates using allograft bone or original autoclaved bone with tumor. Union was complete after 4.6 months, a shorter time than cited in previous reports. ^3,23 The current authors assume that such a high union rate and short union time might be attributable to a smaller reduction in bone morphogenetic protein activity as a result of low heat (65°C for 30 minutes). The union time of the femur was slightly shorter than that of the humerus, although not to a statistically significant degree. Furthermore, as the authors thought that the femur was capable of bearing weight, a compression force was applied which might facilitate healing. However, a distraction force was applied on the humerus, which resulted in a significant rate of nonunion. To promote union, internal fixation was done using a plate and screws and grafted autogenous bone at the junction of the low heat-treated replanted bone and the host bone. The grafted bone was harvested from the medullary canal of the host bone. The plate and screw fixation had possible compression at the osteotomy site, whereas intramedullary long-stem fixation might cause distraction, and possible maintenance of good rotational control, whereas intramedullary long-stem fixation had poor rotational control.

There were no local recurrences in any patients in the current series. Therefore, low heat treatment at 65°C for 30 minutes seems to be sufficient for tumor cell necrosis and a safe method for limb salvage. Reimplantation of original low heat-treated tumor-bearing bone had a similar rate of local recurrence to reimplantation of original autoclaved bone with tumor or allograft bone. The incidence of local recurrence was more strongly related to the status of the surgical margin, as Bacci et al ¹ reported. However, they found that reimplantation of original low heat-treated tumorbearing bone had superior union to that of autoclaved bone or allograft bone.

The scope of the current study was confined to the proximal femur and proximal humerus. Therefore, hips and shoulders were reconstructed using conventional prostheses, consisting of three total hip arthroplasties (Fig 1), three bipolar hemiarthroplasties, and six Neer hemiarthroplasties (Fig 2). Extraarticular wide excision was done including the acetabulum in three of the six patients with proximal femoral lesions, therefore, total hip arthroplasties were done in three patients (Fig 1).

Bone cement was used to fill the medullary cavity of the original low heat-treated tumor-bearing bone to increase the biomechanical property. However, there were two fractures of the reimplanted original low heat-treated tumor-bearing bone; one was the humerus and the other was the femur, both occurring because of a fall. The cause of fractures probably was attributable to a technical error. The humerus fracture occurred at the junction of the stem of the Neer prosthesis and the plate where the stress was concentrated. To avoid this stress concentration, plate fixation should have been extended more proximally. The femur fracture was caused by insufficient fixation. The most proximal hole failed to hold with a screw, and therefore stable fixation was not obtained.

Absorption of original low heat-treated reimplanted tumor-bearing bone occurred in four of the six patients with humeral lesions. However, no absorption was seen in the patients with femoral lesions. This could be because compressive force was applied to the femur, and therefore absorption was prevented. Even though absorption was seen frequently in the humerus, stability was maintained because the medullary canal was filled with bone cement. Loss of stability by massive bone absorption was seen in only one patient and required revision surgery.

The mean survival of original low heat-treated tumor-bearing bone was 57.3 months, median survival was 62 months, and the 5-year survival rate was 83.3%, which were comparable with the survival times reported for endoprostheses. ^9,10,22,24

The functional outcome for the lower extremity was 76.7% (22.8 points), which was comparable or superior to that of endoprosthesis, or allograft bone. ⁶ This is because of anatomic restoration of the lower extremity made possible by using the patient’s bone combined with a conventional prosthesis. However, the functional outcome of the upper extremity was 56.8% (17.0 points), which was relatively poor compared with that of the lower extremity. This was because of the limitation of motion in the shoulder. For a wide margin in the proximal humerus, excision of the rotator cuff was mandatory in most cases.

The 5-year survival rate of the patients was 43.8%, which was lower than rates reported in other studies. ^10,16,25 One patient who had distant metastasis to the lung 17 months after limb salvage surgery died 52 months after surgery. This patient had resection of a metastatic lesion of the lung and received additional chemotherapy. With this patient excluded, the 5-year survival rate would be 87.5%, which is superior to rates reported in other studies. ^10,16,25 The response to preoperative chemotherapy was strongly related to the survival rate of patients. The 5-year survival rate of patients in the poor response group was 0%, but all patients (100%) in the good response group were alive.

Limb salvage with original low heat-treated tumor-bearing bone and a conventional prosthesis might be a useful treatment option for primary bone sarcoma. Based on the results of the current study, limb salvage with original low heat-treated tumor-bearing bone could replace limb salvage with allograft bone, which carries high risks of disease transmission, nonunion, delayed union, and/or immune reaction.